Whole Body Control

Enable easy and safe commanding of the robot’s joints by taking into account its state. Allows for control in joint space, cartesian control of the end-effector, gaze control and much more!

Discover more

Updates Functionalities WBC on TIAGo Demonstrations Robots

The humanoid robot TALOS with the whole body control unit

Discover & define

Whole Body Control (WBC) is a controller written in the ros_control framework developed by PAL Robotics that allows the user to solve a problem consisting of tasks and constraints ordered by priority.

WBC allows the user to add extra functionalities like online teleoperation and dynamic goals for HRI and manipulation that can be combined with Moveit!

The term Whole Body Control refers to the fact that it can command all joints in the robot. This is used, in particular, in all PAL Robotics’ bipeds such as REEM-C, TALOS, and Kangaroo.

What can it do?

The WBC controls all the robot’s joints except for the robot wheels. Hence you will be able to:

  • Control all the joints directly with joint state commands, or maybe a subset of them.
  • Control the joints based on concrete physical input like the head gaze control ,or the position and orientation of the end-effector.
  • Combine the controller with other information shared through ROS .
TIAGo and TIAGo Base
REEM-C

What is it?

Whole Body Control is a feedback optimisation solver written in the ros_control framework.

  • Feedback on commands: WBC uses the current state as input to evaluate commands before performing them like tracking the current position before moving further or not.
  • Optimisation of solution: The controller finds an optimal solution obtained from a mathematical problem that has more conditions than variables.
  • Solve multiple tasks simultaneously: Solves multiple tasks by projecting to the null space of the previous task to force a hierarchical solution

What is it?

Whole Body Control is a feedback optimisation solver written in the ros_control framework. 

  • Feedback: WBC uses the current state as input to evaluate commands before performing them like tracking the current position before moving further or not
  • Optimisation: The controller finds an optimal solution obtained from a mathematical problem that has more conditions than variables.
  • Solve multiple tasks simultaneously: Solves multiple tasks by projecting to the null space of the previous task to force a hierarchical solution.
REEM-C

How to extend these capabilities

Software

You can write your own tasks! PAL Robotics provides tools with commented examples on how to write your own tasks. There are also available tools to manage tasks by: adding and removing tasks to the stack and changing their priorities.

Hardware

You can add external devices to the robot, and interface them with the WBC through ROS. Example of this includes the usage of VR sets for teleoperation and Force/Torque sensors for admittance control.

Whole Body Control in action

Whole Body Control on TIAGo

The default list of tasks for TIAGo robot or stack is composed by 3 actions:

Joint Limit Avoidance

Each joint can exclusively move in the range provided in the robot description, avoiding the limits.

Self Collision Avoidance

The robot uses its own meshes and state to check in real time whether it is in collision with itself to avoid it.

Go-to Default Reference

By default the robot goes to a predefined position to make the problem initially feasible.

Demonstrations

  • Arm teleoperation with a joystick and adding force information to the controller stack

  • Teleoperation of the robot’s upper body by tracking a human hand

  • Aruco demo on how to make the end-effector follow dynamically a marker

  • Rviz interactive markers to move the arm with a digital twin of the robot.

Robots with the Whole Body Control

REEM-C

Bring your research one step ahead with REEM-C. Use the robot to implement and test your algorithms, or to get started in the field of biped robotics.

KANGAROO

The dynamic bipedal platform to research control methods for legged locomotion with a wide range of motion and motor control technology.

TIAGo

The TIAGo robot is ready to deliver social interaction and customized services in a wide variety of sectors. Find a new way to deliver services!

TALOS

The 1.75 m tall humanoid TALOS has sensored torque control at joint level and integrates the latest cutting-edge humanoid technology.

REEM-C
KANGAROO
TIAGo
TALOS

REEM-C

Reem-C, robot produced by PAL Robotics

Bring your research one step ahead with REEM-C. Use the robot to implement and test your algorithms, or to get started in the field of biped robotics.

REEM-C

KANGAROO

The highly dynamic bipedal platform to research advanced control methods for legged locomotion with a wide range of motion and motor control technology.

KANGAROO

TIAGo

Frontal view of TIAGo, the mobile manipulator robot

The TIAGo robot is ready to deliver social interaction and customized services in a wide variety of sectors. Find a new way to deliver services!

TIAGo

TALOS

TALOS, research robot by PAL Robotics

The 1.75 m tall humanoid TALOS has sensored torque control at joint level and integrates the latest cutting-edge humanoid technology.

TALOS

Contact PAL Robotics

Get in touch with us to discover how to use our platforms to engage with people and bring your research or company to the next level.

Contact us